Send metrics on request start

lib/metrics.py

@@ -45,6 +45,7 @@ class Metrics:
         self.model_metrics.workload_received += workload
         self.model_metrics.requests_recieved.add(reqnum)
         self.model_metrics.requests_working.add(reqnum)
+        self.update_pending = True
 
     def _request_end(self, workload: float, reqnum: int) -> None:
         """

workers/comfyui-json/server.py

@@ -33,39 +33,33 @@ log = logging.getLogger(__file__)
 async def generate_client_response(
         client_request: web.Request, model_response: ClientResponse
     ) -> Union[web.Response, web.StreamResponse]:
-        match model_response.status:
-            case 200:
-                log.debug("SUCCESS")
-                # Check if the response is actually streaming based on response headers/content-type
-                is_streaming_response = (
-                    model_response.content_type == "text/event-stream"
-                    or model_response.content_type == "application/x-ndjson"
-                    or model_response.headers.get("Transfer-Encoding") == "chunked"
-                    or "stream" in model_response.content_type.lower()
-                )
+        # Check if the response is actually streaming based on response headers/content-type
+        is_streaming_response = (
+            model_response.content_type == "text/event-stream"
+            or model_response.content_type == "application/x-ndjson"
+            or model_response.headers.get("Transfer-Encoding") == "chunked"
+            or "stream" in model_response.content_type.lower()
+        )
 
-                if is_streaming_response:
-                    log.debug("Detected streaming response...")
-                    res = web.StreamResponse()
-                    res.content_type = model_response.content_type
-                    await res.prepare(client_request)
-                    async for chunk in model_response.content:
-                        await res.write(chunk)
-                    await res.write_eof()
-                    log.debug("Done streaming response")
-                    return res
-                else:
-                    log.debug("Detected non-streaming response...")
-                    content = await model_response.read()
-                    return web.Response(
-                        body=content,
-                        status=model_response.status,
-                        content_type=model_response.content_type
-                    )
-            case code:
-                log.debug(f"Model responded with error {code}")
-                return web.Response(status=code)
-                
+        if is_streaming_response:
+            log.debug("Detected streaming response...")
+            res = web.StreamResponse()
+            res.content_type = model_response.content_type
+            await res.prepare(client_request)
+            async for chunk in model_response.content:
+                await res.write(chunk)
+            await res.write_eof()
+            log.debug("Done streaming response")
+            return res
+        else:
+            log.debug("Detected non-streaming response...")
+            content = await model_response.read()
+            return web.Response(
+                body=content,
+                status=model_response.status,
+                content_type=model_response.content_type
+            )
+            
 
 @dataclasses.dataclass
 class ComfyWorkflowHandler(EndpointHandler[ComfyWorkflowData]):